SOEPpaperson Multidisciplinary Panel Data Research

Stepping Forward: Personality Traits, Choice of Profession, and the Decision to Become Self-Employed

Michael Fritsch and Alina Sorgner

539 201

3SOEP — The German Socio-Economic Panel Study at DIW Berlin 539-2013

SOEPpapers on Multidisciplinary Panel Data Research at DIW Berlin This series presents research findings based either directly on data from the German Socio-Economic Panel Study (SOEP) or using SOEP data as part of an internationally comparable data set (e.g. CNEF, ECHP, LIS, LWS, CHER/PACO). SOEP is a truly multidisciplinary household panel study covering a wide range of social and behavioral sciences: economics, sociology, psychology, survey methodology, econometrics and applied statistics, educational science, political science, public health, behavioral genetics, demography, geography, and sport science. The decision to publish a submission in SOEPpapers is made by a board of editors chosen by the DIW Berlin to represent the wide range of disciplines covered by SOEP. There is no external referee process and papers are either accepted or rejected without revision. Papers appear in this series as works in progress and may also appear elsewhere. They often represent preliminary studies and are circulated to encourage discussion. Citation of such a paper should account for its provisional character. A revised version may be requested from the author directly. Any opinions expressed in this series are those of the author(s) and not those of DIW Berlin. Research disseminated by DIW Berlin may include views on public policy issues, but the institute itself takes no institutional policy positions. The SOEPpapers are available at http://www.diw.de/soeppapers Editors: Jürgen Schupp (Sociology, Vice Dean DIW Graduate Center) Gert G. Wagner (Social Sciences) Conchita D’Ambrosio (Public Economics) Denis Gerstorf (Psychology, DIW Research Director) Elke Holst (Gender Studies, DIW Research Director) Frauke Kreuter (Survey Methodology, DIW Research Professor) Martin Kroh (Political Science and Survey Methodology) Frieder R. Lang (Psychology, DIW Research Professor) Henning Lohmann (Sociology, DIW Research Professor) Jörg-Peter Schräpler (Survey Methodology, DIW Research Professor) Thomas Siedler (Empirical Economics) C. Katharina Spieß (Empirical Economics and Educational Science)

ISSN: 1864-6689 (online)

German Socio-Economic Panel Study (SOEP) DIW Berlin Mohrenstrasse 58 10117 Berlin, Germany Contact: Uta Rahmann | soeppapers@diw.de

Stepping Forward: Personality Traits, Choice of Profession, and the

Decision to Become Self-Employed

Michael Fritsch

Alina Sorgner

January 2013

Abstract

We argue that entrepreneurial choice proceeds in at least in two steps, with vocational choice nearly always preceding choice of employment status, whether that be self-employment or dependent employment. Since the two decisions are interrelated, analysis of entrepreneurial choice as a single act may lead to inconsistent estimates of the factors that determine the decision to launch a business venture. Our empirical analysis utilizes a bivariate probit model that jointly estimates both decisions. The results support our argument that entrepreneurial choice is a two-stage decision process.

JEL classification: L26, J24, D01

Keywords: Entrepreneurial choice, vocational choice, personality traits

Address for correspondence: Michael Fritsch Alina Sorgner Friedrich Schiller University Jena School of Economics and Business Administration Chair of Business Dynamics, Innovation, and Economic Change Carl-Zeiss-Str. 3, D-07743 Jena, Germany m.fritsch@uni-jena.de alina.sorgner@uni-jena.de

1

1. Introduction

The most common model of entrepreneurial choice (Knight, 1921;

Lucas, 1978; Kihlstrom and Laffont, 1979) considers the decision to

start an own firm as a single act and therefore does not adequately

account for the dynamic character of entrepreneurship. However, the

ability and willingness to start an own business do not just happen

overnight, but develop over a period of time based on individual

disposition, qualifications, experience, and, maybe, necessity, such as

lack of an alternative source of income (Parker, 2009; Schmitt-

Rodermund, 2007; Obschonka, Silbereisen and Schmitt-Rodermund,

2010). The common pattern is that people first train for and work in a

certain occupation before they decide to become self-employed

(Fritsch, Kritikos and Rusakova, 2012). Accordingly, the decision to

start an own firm may be partly the result of an individual’s previous

career choices. Due to such interrelatedness with previous decisions,

analysis of entrepreneurial choice as a single act may lead to

inconsistent estimates of the factors that determine the decision to

found an own business.

In this paper we argue that entrepreneurial choice is a two-stage

process, in which the choice of a certain profession made early in life

significantly influences the decision to become self-employed in the

future. Specifically, we discuss possible reasons for interrelatedness

between the choice of a certain profession and the decision to be self-

employed, stressing two main arguments in favor of such

interrelatedness. The first (self-selection hypothesis) states that people

with pronounced entrepreneurial attitudes tend to self-select into certain

professions that might foster development of their entrepreneurial

talent. The second hypothesis (endowment hypothesis) conjectures that

the characteristics of an occupational environment may stimulate the

development of entrepreneurial skills and attitudes even in people

whose personality is not inherently entrepreneurial. In our empirical

2

setting we apply a bivariate probit model to representative data from the

German Socio-Economic Panel (SOEP) and find support for both

hypotheses, suggesting that the choice of a certain occupation and the

decision to become self-employed are indeed interrelated.

The paper proceeds as follows. Section 2 sets out our basic

argument concerning the relationship between vocational choice and

entrepreneurial choice. Section 3 introduces the instrumental variable

approach that allows us to estimate the two interrelated choices.

Section 4 introduces the data and methodology. The results of the

analysis are presented in Section 5. Section 6 concludes and discusses

avenues for further research.

2. Entrepreneurial choice as a two-stage decision process

Chronologically, choosing a profession (vocational choice) nearly

always occurs before deciding to set up an own business. The

relationship between vocational choice and entrepreneurial choice may

have two main forms. First, pro-entrepreneurial people with certain

personality traits such as the ability to bear risk, openness to

experience, and so forth (Caliendo, Fossen and Kritikos, 2009; Zhao

and Seibert, 2006; Zhao, Seibert and Lumpkin, 2010; Rauch and Frese,

2007; Sorgner, 2012) may choose a profession that requires and

rewards entrepreneurial values and provides rich opportunities for self-

employment. Second, individuals who do not choose their profession

with any intention of setting up an own business may be stimulated

toward entrepreneurship while training for and working in such a

profession. This kind of effect may occur through the acquisition of

entrepreneurial skills, contact with entrepreneurial role models,

becoming embedded in a professional network, becoming aware of

entrepreneurial opportunities, or by experiencing poor employment and

earning opportunities in dependent employment (Sorgner and Fritsch,

2013). Figure 1 shows the interplay of different factors that may

influence both the choice of a profession and entrepreneurial choice.

3

1

Step I:Choice of profession

Step II:

Start-up decision

Occupation-specific

human capital

● Formal education

● Structure of skill

portfolio

Occupation-specific environment:

● Conditions in the labor market

● Entrepreneurial career models

● Minimum efficient size

● Entry regulation

Individual characteristics:

● Personality traits

● Career role models

● Socio-demographic

characteristics

Figure 1: Entrepreneurial choice as a two-stage decision process.

Factors often regarded as motivation to launch an own business,

such as wanting to earn more money, a wish for more independence

and flexibility, and a desire for more creative work, among others, may

also guide vocational choice. Hence, people with an entrepreneurial

mindset may be likely to choose occupations that match their

preferences. For instance, someone with a strong motivation to earn

money will be more likely to enter an occupation that offers relatively

high earnings (Sorgner and Fritsch, 2013). Similarly, people who enjoy

independence and a creative work environment may choose an artistic

occupation. And, indeed, research shows that people with

entrepreneurial personality traits (e.g., high levels of openness to

experience, conscientiousness, extraversion, and low degrees of

neuroticism and agreeableness) are more likely to be interested in so-

4

called enterprising occupations, such as management, sales, and

finance, and are relatively often found in artistic occupations (e.g.,

musician, artist, architect) (Schmitt-Rodermund, 2004; Sorgner, 2012).

The fact that the majority of start-ups occur after the founder has

spent some time in dependent employment (Fritsch, Kritikos and

Rusakova, 2012) suggests that occupational environment may have a

considerable impact on an individual’s decision to become self-

employed (Sorgner and Fritsch, 2013). One aspect of occupational

environment that might have such an influence is the variety of skills

required. According to the “jack of all trades” hypothesis (Lazear, 2004,

2005), work environments characterized by high task complexity and

that require a great number of different skills should be especially

conducive to self-employment. Indeed, Fritsch, Bublitz and Rusakova

(2012) show that occupational environments vary greatly with regard to

the skill balance of the people working in them, which may partly

explain differences in start-up rates across occupations. Specifically,

people in occupations that require high levels of human capital (e.g.,

engineers, chemists, physicists, natural scientists, managers) tend to

also have high skill balance, whereas unskilled workers, such as street

vendors, cleaners, or delivery service drivers, generally have rather low

skill balance levels.

Occupational environment may also be conducive to an individual’s

decision to become an entrepreneur if a great many entrepreneurs

already work in the occupation, thereby providing role models for this

career path. For instance, people in professions with relatively high self-

employment rates, such as medicine, law, and art, are more often

exposed to examples of an entrepreneurial career in their profession as

compared to people in professions characterized by a lower level of

self-employment. Observing self-employed peers and colleagues is

found to have a strong effect on an individual’s propensity to become

self-employed (Nanda and Sorensen, 2010; Bosma et al., 2012). The

presence of established entrepreneurial businesses in many of the

5

liberal professions, for example, self-employed physicians, lawyers, and

tax consultants, are recognized and relatively easily reproducible

models. Moreover, such standardized entrepreneurial opportunities may

be easier to “sell” to financiers. Furthermore, entry regulations in certain

professions, such as physicians, psychotherapists, and lawyers, may

result in a considerably lower risk of failure when starting an own firm.

Indeed, people who work in occupations characterized by high levels of

self-employment are more likely to become self-employed themselves

(Sorgner and Fritsch, 2013). Another characteristic of many professions

with high levels of self-employment, and one that may be especially

conducive to starting an own business, is small minimum efficient size,

thus requiring low initial investment (Figure 1).

Poor economic conditions in the occupation-specific labor market,

such as scarcity of jobs, low income in dependent employment, and

high levels of unemployment, may influence a person’s entrepreneurial

choice even if he or she had no original intentions of becoming self-

employed. It is plausible to assume that many people are well aware of

the opportunities available in the occupation-specific labor market when

making their vocational choice. During training for and working in a

chosen occupation, people acquire specific human capital that is often

hardly transferable across occupations (Gathmann and Schönberg,

2010). Hence, if economic conditions in the occupation-specific labor

market change, the costs associated with re-qualification might prevent

people from switching to another occupation with different qualification

requirements (Nedelkoska and Neffke, 2010), making a switch into self-

employment in the given occupation relatively attractive. Sorgner and

Fritsch (2013) indeed find that occupation-specific unemployment rates

are positively associated with the propensity to become self-employed.

All in all, there are numerous indications that entrepreneurial choice

does not occur in one stage, but that it involves at least two decisions,

including the choice of profession. The choice of profession and the

decision to start an own business may be interrelated in two ways. First,

6

people with a strong desire to become self-employed may choose

professions that make it more feasible to set up an own business.

Second, people may first choose a profession that matches their

interests, personalities, and preferences independent of any intention of

future self-employment but then, later on, based on the experience they

gain in this profession, decide to become self-employed. Hence, there

are good reasons to expect that there is a causal relationship between

professional and entrepreneurial choices.

In this dynamic process, the choice of a profession may be viewed

as an advance decision for or against self-employment. If, however, the

decisions are interrelated, analysis of entrepreneurial choice as a single

act may lead to inconsistent estimates of the factors that determine the

decision to launch a business venture. Hence, a correct estimation of

the model of entrepreneurial choice needs to account for the choice of a

profession.

3. The effect of the choice of profession on entrepreneurial choice: An instrumental variable approach

The arguments presented above suggest that there is a causal

relationship between professional and entrepreneurial choice. However,

trying to account for such a two-stage entrepreneurial choice by simply

including profession as an independent variable in an empirical model

may not solve the problem since there could be unobservable

characteristics that are correlated with both the choice of profession and

entrepreneurial choice. To test the relevance of this concern, we apply

an instrumental variables approach that should allow us to identify the

causal effect. We instrument an individual’s choice of profession with

that practiced by his or her father when the individual was about 15

years old. A causal interpretation of the relationship between the choice

of a profession and entrepreneurial choice requires that the instrument

is relevant and independent of the outcome variable (respondent’s

entrepreneurial choice).

7

An instrument is appropriate if it is correlated with the endogenous

variable, in our case, the respondent’s choice of a profession. There are

numerous studies finding a pronounced tendency for intergenerational

occupation following, that is, children choose the same occupation as

their parents (Jonsson et al., 2009; Laband and Lentz, 1983, 1989;

Karlsson and Stanfors, 2011). Laband and Lentz (1983) argue that a

main reason for occupation following is the transfer of occupation-

related skills and reputation from parents to children. Since occupation

is an important part of person’s identity, parents might foster the taste

for a certain occupation in their children. A cross-national study by

Jonsson et al. (2009) finds that occupational inheritance is particularly

strong in Germany, which may be due to the well-developed system of

vocational training in this country, leading to a considerable

accumulation of occupation-specific skills and thus to a strong

commitment to one’s occupation and a willingness to transfer one’s

occupational identity to one’s children. Occupation following may also

result from the selection processes of training institutions given that

these sometimes favor students who come from families already active

in the chosen profession. Indeed, Laband and Lentz (1989) show that

children of medical doctors are more likely to be admitted to medical

school compared to children of parents in non-medical professions, and

this was discovered to be at least partly due to personal networks and

nepotism.

Another necessary condition for an instrument variable is its

independence of the outcome variable. Empirical evidence suggests

that a child’s self-employment status is independent of the father’s

occupation. In particular, Dunn and Holtz-Eakin (2000, 290) show that

self-employed sons were as likely as never-self-employed sons to

follow their father’s occupation. Moreover, the authors reveal that the

majority of self-employed sons in their data entered occupations others

than those of their (self-employed or non-self-employed) fathers.

Additionally, those people who do not follow in their parents’

8

occupational footsteps often use self-employment as a way of moving

up the socioeconomic status ladder (Katz, 1992). Finally, unwillingness

to follow the father’s occupation may be related to rule-breaking

behavior in adolescence, which involves defying parental authority and

expectations. Rule breaking in adolescence, in turn, is shown to be

positively associated with entrepreneurial status (Zhang and Arvey,

2009). Thus, it appears plausible that occupation followers are not more

likely to be self-employed than non-occupation followers. Overall, our

instrument appears to fulfill the necessary conditions. We provide

additional proof of the independence of our outcome variable from the

instrument in Section 5.1.

4. Data and methodology

4.1 Sample and measures

Our empirical analysis is based on the German Socio-Economic Panel

(SOEP), a representative longitudinal study of private households in

Germany (for details, see Wagner, Frick and Schupp, 2007). For the

present analysis, we use the 2009 wave of the survey because the full

set of independent variables, including information on personality

characteristics, was available in this particular year. The 2009 wave of

the SOEP provides data on about 21,000 individuals living in Germany.

We exclude persons who were retired, unemployed, or engaged in full-

time education. We also exclude civil servants and those in military

service since we consider the choice of profession for these groups to

be different from that of employees in the private sector. Self-employed

farmers and employees in the agricultural sector are excluded for the

same reason.1 Next, all persons who stated that their primary activity is

helping in a family business are omitted because of their mixed status

of being neither “full” entrepreneurs nor “pure” dependent employees.

1 Most farms in Germany are family businesses so that their owners are more or less self-employed by definition. Thus, the self-employment of farmers may be a result of a either family or regional tradition.

9

After excluding respondents with missing values for relevant

information, our sample includes 5,100 persons, 630 (12.4 percent) of

whom are self-employed in their main occupation.

A key variable in our analysis is an individual’s profession. The

information on individuals’ professions is available at the four-digit level

of the International Classification of Occupations (ISCO-88; for details,

see International Labour Office, 1990). To have a sufficiently large

number of cases for analysis, we create eight groups of professions that

correspond to the two-digit level of ISCO-88 (Table 1). In this

classification scheme, occupations are aggregated on the basis of

similarity of skills required to fulfill the tasks and duties of the jobs and

not according to the level of professional hierarchy that individuals

achieved. Hence, older persons who may have achieved a high

position, such as director, are in the same category as beginners in that

Table 1: Sample occupations

Occupational group:

2-digit ISCO-88

codes Description

Managers 11-13 Directors, chief executives, department managers, general managers

Physical, mathematical, and engineering science professionals

21, 31 Physicists, mathematicians, computing professionals, architects, engineers, technicians, and associate professionals

Life science and health professionals

22, 32 Medical doctors, veterinarians, pharmacists, biologists

Teaching professionals 23, 33 College, university, and higher education teaching (associate) professionals

Other professionals 24, 34 Business professionals, legal professionals, social science professionals, writers, and creative or performing artists

Service workers 51, 52 Travel attendants, housekeepers, cooks, personal care workers, hairdressers, protective service workers

Trades workers 71-74 Roofers, painters, miners, builders, musical instrument makers, bookbinders, shoemakers

Elementary occupations 41, 42, 81-

93

Office clerks, customer services clerks, machine operators and assemblers, drivers, sales and service elementary occupations

10

particular profession. The same occupational classification scheme is

used to identify the profession of a respondent’s father when the

respondent was 15 years old.

Previous empirical analyses of the determinants of self-employment

find a significant impact of various human and social capital, socio-

demographic characteristics, and the macro environment on the

probability of running an own business (see Parker, 2009). In our

model, we account for these influences to the extent that appropriate

indicators are available in the data. The set of available control

variables includes age, gender, marital status, nationality, years of

formal education, years of unemployment, parental role models of self-

employment at age of 15, and region in which a person resides. Since

the 2009 wave of the SOEP data contains questions on the Big Five

dimensions of personality, we are able to control for personality traits.2

Furthermore, we include a measure of risk preferences in our analysis,

since entrepreneurs are commonly regarded as risk-takers (Caliendo,

Fossen and Kritikos, 2009). The measure of risk attitudes in SOEP is an

experimentally validated measure based on the question “Are you

generally a person who is fully prepared to take risks or do you try to

avoid taking risks?” (Dohmen et al., 2011). A description of all variables

used in the analysis, as well as the correlation matrix, can be found in

Tables A1 and A2 in the Appendix.

4.2 Methodology: The bivariate probit model

In the empirical analysis, we employ a bivariate probit model (Greene,

2008) for a joint estimation of the two interrelated choices—choice of

2 These questions refer to a psychological scale that measures the Big Five factors (Costa and McCrae, 1992) based on three questions for each of the broad dimensions. The SOEP respondents were asked to rate themselves on a seven-point scale, with 1 indicating that a given personality characteristic does not apply to them at all and 7 meaning that the characteristic applies perfectly. The value for each of the Big Five dimensions is an arithmetical mean of the responses to the related three questions. A detailed description of the procedure used in the SOEP survey can be found in Gerlitz and Schupp (2005). The authors show that self-reported personal attitudes based on the Big Five related questions in the SOEP are valid and reliable.

11

profession and decision to be self-employed—in order to obtain

consistent estimates of coefficients.3

Consider the model,

,

,

222

)(*

2

1

)(

2

)(

11

*

1

εβ

εγβ

+′=

++′=

xy

yxy

k

kk

(1)

where *

1y and )*(

2

ky are latent variables. We observe:

>=

>=

.0,0,1

,0,0,1

)(*

2

)(

2

*

11

otherwiseyify

otherwiseyify

kk

The dependent variable, 1y , is equal to 1 if the individual decides to be

self-employed; 0 otherwise. The variable,)(

2

ky , assumes the value 1 if an

individual chooses profession k , Nk ,1∈ , and 0 otherwise; 1x and

2x are vectors of explanatory variables, 1β ′ and 2β ′ are parameter

vectors. We are interested in estimating the scalar parameter )(kγ ,

which indicates the impact of the predicted probability of an individual to

be in profession k on the likelihood of being self-employed. The error

terms 1ε and 2ε are assumed to be jointly normally distributed with

,),|,(

,1),|(),|(

,0),|(),|(

2121

212211

212211

ρεε

εε

εε

=

==

=Ε=Ε

xxCorr

xxVarxxVar

xxxx

(2)

and the joint density of the error terms is given by

3 A modification of a bivariate model, a recursive bivariate probit model, is discussed and applied empirically in several studies (see, e.g., Greene, 1998, 2008; Bauernschuster, Falck, and Heblich, 2009; Fabbri and Monfardini, 2008; Kassouf and Hoffmann, 2006; Rhine, Greene, and Toussaint-Comeau, 2006).

12

.1

2

2

1exp

12

1),(

2

21

2

2

2

1

2

2

21

1

−

−+−

−=

ρ

ερεεε

ρσπσεεφ

εε

(3)

If 0=ρ , the equations in (1) are not interrelated and thus can be

estimated separately using a univariate probit model. If 0≠ρ , the joint

estimation of both equations using the bivariate probit model will

provide consistent parameter estimates.

5. Results

5.1 Descriptive statistics

We find a number of statistically significant differences between the

self-employed and the reference group of dependently employed

individuals in our sample (Table 2).4 With regard to regional

environment, we find that the average propensity for self-employment is

significantly higher in large agglomerations and significantly lower in

urbanized regions. However, there is no statistically significant

difference between self-employed and dependently employed people in

rural areas. On average, self-employed persons had 14.1 years of

education, which is significantly more than the average 12.7 years of

education of the dependently employed. Moreover, about 19.2 percent

of self-employed persons had self-employed parents when they were

15 years old; this figure is 8.9 percent for the dependently employed.

Self-employed individuals in the sample are more likely to be married

(68.7 percent) than are employees (60.3 percent) and there is a

significantly higher share of males in the group of self-employed (64.6

percent) than in the group of dependently employed (50.7 percent). The

average age of self-employed persons in the sample is about 48.2

years old, which is about 6 years more than the average age of

employees.

4 For descriptive statistics of all variables, see Table A2 in the Appendix.

13

Table 2: Mean characteristics and t-test of equal means for self-employed individuals compared to the reference group

Variable: Self-employed Dependently

employed

Agglomerations 0.565*** 0.492

Urbanized regions 0.321*** 0.377

Rural areas 0,114 0.131

Years of education 14.084*** 12.699

Years unemployed 0.496* 0.618

Either parent self-employed (yes = 1, no = 0) 0.192*** 0.089

Married (yes = 1, no = 0) 0.687*** 0.603

Male (yes = 1, no = 0) 0.646*** 0.507

German (yes = 1, no = 0) 0,960 0.953

Age 48.195*** 42.274

Openness 4.854*** 4.417

Conscientiousness 5.964** 5.882

Extraversion 5.012*** 4.817

Agreeableness 5,275 5.257

Neuroticism 3.506*** 3.739

Willingness to take risks 4.557*** 4.040

Number of observations 630 4,470

Notes: ***: statistically significant at the 1% level; **: statistically significant at the 5% level; *: statistically significant at the 10% level.

As Figure 2 shows, self-employment is not equally distributed

across the different groups of professions. The highest average

probabilities of self-employment are in professions such as

management (30.8 percent), teaching (18.1 percent), and life science

and health (17.8 percent). The lowest average probabilities of self-

employment are in elementary occupations such as office and customer

service clerks (2.7 percent) and service workers and sales (7.5

percent).

14

0.308

0.181

0.178

0.164

0.129

0.115

0.075

0.027

0 .1 .2 .3Share of self-employed

Managers

Teaching professionals

Life science and health professionals

Other professionals

Trades workers

Phys., math. and engin. professionals

Services workers and salespersons

Elementary occupations

Figure 2: Shares of self-employed across occupations

Table 3: Occupational role models

Number of respondents in

occupation Father was in the same occupation Share, %

Managers 364 63 17,31%

Physicists, mathematicians, and engineering professionals 635 113 17,80%

Life science and health professionals 342 25 7,31%

Teaching professionals 249 19 7,63%

Other professionals 1.116 206 18,46%

Service workers 549 20 3,64%

Trades workers 641 301 46,96%

Elementary occupations 1.204 384 31,89%

Total 5100 1131 22,18%

15

With regard to the relevance of our instrument variable, we find that

22.2 percent of the individuals in our sample have followed the

occupational pattern of their fathers (Table 3). However, occupation

followers are unevenly distributed across the groups of professions. The

highest share (almost 47 percent) of those choosing their father’s

profession is in the group of trades workers. High shares of occupation

followers are also observed among elementary occupations (32

percent) that involve simple and routine tasks and in the group of “other

professionals” (18.5 percent). The lowest share of occupation followers

is among service workers (3.6 percent). This relatively low share of

intergenerational transmission among service workers may be partly

explained by the recent expansion of the service sector, which may

have offered employment opportunities to a wide spectrum of the

population, not only those whose fathers worked in the service sector.

Proving the instrumental variable’s independence from the outcome

variable is much more challenging than demonstrating the relevance of

the instrument. For instance, one possible concern about our instrument

variable is that occupation followers of self-employed fathers may be

more likely to be self-employed themselves because they might be

qualified enough to take over a family business or just because they are

more prone to have the same occupational tastes as their fathers,

including the attitude toward self-employment. Hence, to gain some

insight into whether occupation followers are more likely to be self-

employed we conduct a probit regression analysis (Table 4). The binary

dependent variable indicates whether a person has chosen the

profession her or his father had when the person was 15 years old (yes

= 1, no = 0), and the independent variables are the child’s self-

employment status (self-employed vs. employee), father’s self-

employment status when the respondent was about 15 years old, and a

set of control variables, including personality traits. The results suggest

that neither the self-employment status of a child nor the self-

16

Table 4: Determinants of occupational following

Dependent variable: occupational following (yes = 1, no = 0) Coefficient

Standard error

Self-employed (yes = 1, no = 0) 0.075 (0.062)

Either parent self-employed (yes = 1, no = 0) 0.061 (0.065)

Age 0.000 (0.013)

Age, squared 0.000 (0.000)

Male (yes = 1, no = 0) 0.418*** (0.043)

Married (yes = 1, no = 0) 0.004 (0.044)

Conscientiousness -0.009 (0.024)

Extraversion 0.027 (0.019)

Agreeableness -0.018 (0.021)

Openness -0.033*** (0.019)

Neuroticism 0.018 (0.017)

German nationality (yes = 1, no = 0) -0.131 (0.090)

Years of education -0.025*** (0.008)

Intercept -0.433 (0.360)

Number of observations 5,100

R² 0.025

Wald chi² 124.89***

Log-likelihood -2,630.87

Notes: Results of a probit regression. The dependent variable equals 1 if the respondent has the same occupation that his or her father had when the respondent was 15 years old; 0 otherwise. Robust standard errors in parentheses. ***: statistically significant at the 1% level; **: statistically significant at the 5% level; *: statistically significant at the 10% level.

employment status of her or his parents has a statistically significant

association with the probability of being an occupation follower.

Interestingly, those people with high scores on the “openness to

experience” scale, which is an important trait for an entrepreneur

(Schmitt-Rodermund, 2007; Sorgner, 2012), are significantly less likely

to be occupation followers. Thus, we found no indication that

17

occupation following is associated with the probability of self-

employment and we therefore regard our instrument variable as an

independent one.

5.2 Multivariate analysis

Following the model described in previous section, 1y equals 1 if a

person is self-employed and equals 0 if a person is dependently

employed; )(

2

ky

equals 1 for persons whose profession is k , 8,1∈k and

equals 0 otherwise; 1x is a vector of explanatory variables that

comprises all variables reported in Table 2, i.e., those factors that

appear to have a significant association with the probability of self-

employment.

The literature identifies a number of factors that might be

associated with the choice of profession. For instance, the theory of

vocational behavior (Holland, 1985; Schneider, 1987) argues that

choice of profession is an expression of one’s personality and, indeed,

several empirical studies show a relationship between personality and

choice of a particular vocation (Filer, 1986; Borghans et al., 2008;

Sorgner, 2012). Moreover, certain professions continue to be either

male or female dominated. And, as discussed above, family

occupational role models may have an important influence on a child’s

vocational choice. Thus, 2x , the vector of explanatory variables in the

second equation in (1), includes the Big Five traits, willingness to take

risks, gender, parental self-employment status when the respondent

was 15 years old, and a dummy variable indicating whether a person

has the same profession as his or her father had when the respondent

was 15 years old.

We first estimate a univariate probit model of self-employment that

includes all variables in 1x and a dummy variable for each of the eight

groups of professions (Table 5). The results indicate a statistically

18

significant effect of profession on the probability of self-employment in

six of the eight groups. We find a positive effect of profession on the

probability of being self-employed for managers, life science and health

professionals, trades workers, and other professions. In physical,

mathematical, and engineering science professions, as well as in

elementary occupations, this effect is significantly negative, indicating a

relatively low propensity for self-employment. No statistically significant

relationships are found for teaching professionals or service workers.

As these estimates may be inconsistent due to possible

interdependency of the choice of profession and self-employment

status, the bivariate probit model is estimated in the next step (Table 6).

A first important result of the bivariate probit regression is that ρ is

significantly different from zero in model specifications for managers,

physical, mathematical, and engineering professionals, other

professionals, service workers, and elementary occupations. Thus, the

choice of one of these professions and entrepreneurial choice are

interrelated. However, ρ is not significantly different from zero in

models for life science and health professionals, teachers, and trades

workers. This means that the choice of one of these professions and

self-employment choice are not interrelated, and one must interpret the

corresponding results from the univariate probit model (Table 5).

Concerning our instrument variable, we find that the profession of the

respondent’s father when the respondent was 15 years old strongly

determines the respondent’s choice of profession. This effect is

statistically significant for all professional groups except service

workers. Moreover, the results of the bivariate probit model are quite

different from those of the univariate probit model. For instance, in the

univariate probit model, being a manager had a positive effect on the

probability of self-employment (Table 5), but we now observe that the

predicted probability of being a manager has no significant effect on the

19

Table 5: Results of a univariate probit regression for the probability of self-employment in the full sample with dummy variables indicating a type of profession

I II III IV V VI VII VIII

Variables: k=Managers

k=Physical, mathematical, engineering

science profes-sionals

k=Life science and health

professionals k=Teaching

professionals k=Other profes-

sionals k=Service workers

k=Trades workers

k=Elementary occupations

Dummy: =1 if profession=k 0.526*** -0.319*** 0.221** 0.125 0.161*** -0.016 0.208*** -0.791*** (0.082) (0.076) (0.088) (0.107) (0.057) (0.091) (0.076) (0.088) Agglomerations 0.039 0.049 0.044 0.050 0.047 0.047 0.051 0.0522 (0.078) (0.078) (0.077) (0.077) (0.078) (0.077) (0.078) (0.079) Urbanized regions -0.025 -0.017 -0.025 -0.015 -0.013 -0.018 -0.019 -0.013 (0.080) (0.080) (0.080) (0.080) (0.080) (0.080) (0.081) (0.082) Years of unemployment 0.015 0.008 0.011 0.010 0.012 0.010 0.010 0.029*

(0.015) (0.015) (0.015) (0.015) (0.015) (0.015) (0.015) (0.015) Either parent self-employed (yes = 1, no = 0)

0.369*** 0.374*** 0.378*** 0.385*** 0.384*** 0.383*** 0.378*** 0.359***

(0.070) (0.070) (0.070) (0.070) (0.070) (0.070) (0.070) (0.071) Married (yes = 1, no = 0) 0.055 0.070 0.065 0.068 0.071 0.067 0.067 0.053

(0.055) (0.054) (0.054) (0.054) (0.054) (0.054) (0.054) (0.055) German nationality (yes = 1, no = 0)

-0.114 -0.082 -0.083 -0.071 -0.079 -0.076 -0.072 -0.079

(0.120) (0.120) (0.120) (0.120) (0.121) (0.120) (0.121) (0.125) Age -0.003 0.0018 0.001 0.001 0.003 0.001 0.001 0.003 (0.015) (0.015) (0.015) (0.015) (0.015) (0.015) (0.015) (0.015) Age, squared 0.0003* 0.0003 0.0003* 0.0003* 0.0002 0.0003* 0.0003* 0.0003 (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) Years of education 0.079*** 0.087*** 0.078*** 0.079*** 0.076*** 0.080*** 0.087*** 0.063*** (0.009) (0.009) (0.009) (0.009) (0.009) (0.009) (0.009) (0.009)

20

Table 5 (continued) Male (yes = 1, no = 0) 0.251*** 0.327*** 0.303*** 0.295*** 0.302*** 0.285*** 0.251*** 0.268*** (0.053) (0.054) (0.053) (0.053) (0.053) (0.053) (0.055) (0.054) Openness 0.108*** 0.114*** 0.110*** 0.107*** 0.108*** 0.109*** 0.110*** 0.091*** (0.025) (0.025) (0.025) (0.025) (0.025) (0.025) (0.025) (0.025) Conscientiousness 0.041 0.040 0.037 0.042 0.043 0.040 0.035 0.043 (0.031) (0.031) (0.031) (0.031) (0.031) (0.031) (0.031) (0.031) Extraversion 0.065*** 0.059** 0.071*** 0.069*** 0.066*** 0.069*** 0.069*** 0.075*** (0.024) (0.024) (0.024) (0.024) (0.024) (0.024) (0.024) (0.024) Agreeableness 0.001 -0.003 -0.003 -0.004 -0.0002 -0.002 -0.001 -0.003 (0.027) (0.027) (0.027) (0.027) (0.027) (0.027) (0.027) (0.027) Neuroticism -0.010 -0.014 -0.013 -0.013 -0.012 -0.013 -0.016 -0.008 (0.022) (0.022) (0.022) (0.022) (0.022) (0.022) (0.022) (0.022) Willingness to take risks 0.046*** 0.052*** 0.053*** 0.052*** 0.052*** 0.052*** 0.051*** 0.053*** (0.012) (0.012) (0.012) (0.012) (0.012) (0.012) (0.012) (0.013) Intercept -4.128*** -4.336*** -4.262*** -4.283*** -4.332*** -4.284*** -4.360*** -3.968*** (0.469) (0.469) (0.468) (0.468) (0.470) (0.470) (0.472) (0.481) Observations 5,100 5,100 5,100 5,100 5,100 5,100 5,100 5,100 Log-likelihood -1,660 -1,675 -1,680 -1,683 -1,679 -1,683 -1,680 -1,630 Chi2 472.4*** 431.9*** 418.2*** 422.9*** 419.6*** 418.6*** 418.5*** 434.4*** Pseudo R2 0.129 0.122 0.119 0.118 0.119 0.117 0.119 0.145

Notes: Results of a univariate probit regression. Dependent variable: Self-employment status (yes = 1, no = 0). Robust standard errors in parentheses. ***: statistically significant at the 1% level; **: statistically significant at the 5% level; *: statistically significant at the 10% level.

21

Table 6: Probabilities of self-employment and choice of profession: Bivariate probit model

I II III IV V VI VII VIII

k = Managers

k = Physicists, mathematicians, and

engineering professionals

k = Health professionals

k = Teaching professionals

k = Other professionals k = Service workers k = Trades workers

k = Elementary occupations

Variables:

Self-em-ployment

choice Vocational

choice

Self-em-ployment

choice Vocational

choice

Self-em-ployment choice

Vocational choice

Self-em-ployment

choice Vocational

choice

Self-em-ployment

choice Vocational

choice

Self-em-ployment

choice Vocational

choice

Self-em-ployment

choice Vocational

choice

Self-em-ployment

choice Vocational

choice

Predicted probability of having profession k

-0.0832 (0.227)

- 0.790*** (0.294)

- 0.550 (1.633)

- 0.612 (0.450)

- 1.031*** (0.238)

- 1.044** (0.473)

- -0.171 (0.516)

- -2.007*** (0.083)

-

Father had profession k -

0.333*** (0.081)

- 0.364*** (0.066)

- 0.830*** (0.156)

- 0.572*** (0.142)

- 0.432*** (0.056)

- 0.218

(0.152) -

0.371*** (0.050)

- 0.207*** (0.039)

Either parent self-employed (yes = 1, no = 0)

0.387*** (0.070)

0.129 0.364*** -0.0300 0.368*** 0.0889 0.384*** -0.0205 0.333*** 0.0881 0.371*** 0.0176 0.372*** -0.0655 0.114* -0.257*** (0.084) (0.069) (0.078) (0.087) (0.089) (0.070) (0.103) (0.072) (0.065) (0.069) (0.081) (0.071) (0.083) (0.067) (0.069)

Male (yes = 1, no = 0)

0.288*** 0.420*** 0.220*** 0.681*** 0.322*** -0.549*** 0.316*** -0.526*** 0.358*** -0.332*** 0.371*** -0.764*** 0.328*** 1.145*** 0.130*** -0.095** (0.054) (0.062) (0.056) (0.055) (0.096) (0.061) (0.055) (0.070) (0.052) (0.043) (0.056) (0.055) (0.125) (0.064) (0.047) (0.041)

Openness 0.116*** 0.090*** 0.091*** 0.116*** 0.108*** 0.066** 0.099*** 0.208*** 0.089*** 0.061*** 0.114*** -0.080*** 0.101*** -0.085*** -0.032 -0.161*** (0.025) (0.0255) (0.024) (0.023) (0.029) (0.028) (0.0256) (0.0308) (0.0239) (0.0185) (0.024) (0.022) (0.028) (0.024) (0.024) (0.018) Conscientiousness 0.042 0.017 0.049 -0.052* 0.035 0.055* 0.046 -0.123*** 0.053* -0.062*** 0.040 -0.007 0.046 0.128*** 0.043* 0.024 (0.031) (0.032) (0.030) (0.028) (0.033) (0.033) (0.031) (0.040) (0.0300) (0.023) (0.0303) (0.031) (0.034) (0.030) (0.025) (0.023) Extraversion 0.066*** 0.012 0.081*** -0.152*** 0.075*** -0.096*** 0.069*** -0.015 0.056** 0.032 0.060** 0.074*** 0.070*** 0.019 0.080*** 0.050*** (0.024) (0.026) (0.023) (0.022) (0.028) (0.027) (0.024) (0.029) (0.023) (0.019) (0.023) (0.024) (0.024) (0.023) (0.0193) (0.018) Agreeableness -0.004 -0.049* -0.001 -0.025 -0.004 0.037 -0.008 0.116*** 0.002 -0.015 -0.007 0.051* -0.003 -0.018 -0.0138 -0.011 (0.027) (0.029) (0.026) (0.025) (0.028) (0.029) (0.027) (0.036) (0.026) (0.022) (0.026) (0.028) (0.027) (0.026) (0.022) (0.021) Neuroticism -0.017 -0.075*** -0.008 -0.035* -0.011 -0.041* -0.012 -0.021 -0.003 -0.043** -0.016 0.012 -0.009 0.079*** 0.025 0.063*** (0.022) (0.024) (0.021) (0.021) (0.023) (0.024) (0.022) (0.028) (0.021) (0.017) (0.021) (0.022) (0.024) (0.022) (0.018) (0.017)

22

Table 6 (continued)

Willingness to take risks

0.051*** 0.048*** 0.052*** -0.009 0.053*** -0.021 0.054*** -0.040** 0.051*** -0.008 0.051*** -0.008 0.053*** 0.031** 0.037*** -0.001 (0.012) (0.014) (0.012) (0.0118) (0.013) (0.014) (0.012) (0.016) (0.012) (0.010) (0.012) (0.013) (0.013) (0.012) (0.011) (0.010)

Agglomerations 0.038 - 0.042 - 0.044 - 0.048 - 0.037 - 0.048 - 0.051 - 0.031 - (0.076) (0.074) (0.077) (0.077) (0.074) (0.075) (0.077) (0.053) Urbanized regions -0.022 - -0.018 - -0.024 - -0.017 - -0.016 - -0.019 - -0.016 - -0.014 - (0.079) (0.077) (0.080) (0.079) (0.076) (0.078) (0.079) (0.054) Years of education 0.077*** - 0.081*** - 0.078*** - 0.079*** - 0.071*** - 0.079*** - 0.085*** - 0.043*** -

(0.009) (0.009) (0.010) (0.009) (0.009) (0.009) (0.010) (0.007) Years of unemployment

0.015 - 0.008 - 0.011 - 0.011 - 0.011 - 0.011 - 0.009 - 0.023** - (0.014) (0.014) (0.015) (0.0145) (0.014) (0.014) (0.015) (0.011)

Married (yes = 1, no = 0)

0.054 - 0.064 - 0.066 - 0.0676 - 0.066 - 0.068 - 0.067 - 0.044 - (0.054) (0.052) (0.054) (0.0540) (0.051) (0.053) (0.054) (0.037)

German (yes = 1, no = 0)

-0.111 - -0.067 - -0.082 - -0.0748 - -0.081 - -0.083 - -0.067 - -0.068 - (0.119) (0.115) (0.120) (0.119) (0.115) (0.117) (0.121) (0.084)

Age -0.003 - 0.003 - 0.001 - 0.0004 - 0.003 - 0.0004 - 0.001 - -0.001 - (0.015) (0.014) (0.015) (0.015) (0.014) (0.015) (0.015) (0.010) Age, squared 0.0003* - 0.0002 - 0.0003 - 0.0003* - 0.0002 - 0.0003* - 0.0003* - 0.0002* - (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) Intercept -4.08*** -2.028*** -4.374*** -0.832*** -4.280*** -1.430*** -4.26*** -2.034*** -4.338*** -0.470** -4.272*** -1.189*** -4.380*** -2.865*** -2.19*** -0.550*** (0.468) (0.295) (0.452) (0.247) (0.469) (0.295) (0.468) (0.333) (0.453) (0.208) (0.463) (0.265) (0.469) (0.274) (0.418) (0.206) Observations 5,100 5,100 5,100 5,100 5,100 5,100 5,100 5,100 5,100 5,100 5,100 5,100 5,100 5,100 5,100 5,100 Rho 0.311 0.311 -0.560 -0.560 -0.161 -0.161 -0.230 -0.230 -0.488 -0.488 -0.506 -0.506 0.217 0.217 0.908 0.908 Wald test of rho=0: Wald chi2

6.854*** 11.33*** 0.041 1.323 10.85*** 4.315** 0.500 32.94***

Notes: Robust standard errors in parentheses. ***: statistically significant at the 1% level; **: statistically significant at the 5% level; *: statistically significant at the 10% level.

23

probability of self-employment. This result indicates that those persons

who are more likely to be managers due to their personal

characteristics are not more likely to become self-employed. The

significant and positive coefficient gained in the univariate probit

regression mainly highlights the importance of a managerial

occupational environment for self-employment, for instance, through

developing the entrepreneurial skills necessary for managing an own

business. However, this positive effect is less likely due to an above-

average proclivity of “managerial” types for entrepreneurship; rather,

“managerial” types are on average equally likely to choose self-

employment or dependent employment.

There are two effects that do not change sign, but do become

stronger in the bivariate probit setting as compared to the univariate

model. One of these effects concerns the group of other professionals,

which includes business professionals, legal professionals, social

sciences professionals, and artists. The personal characteristics of

people who are more likely to be in one of these professions imply a

relatively high propensity for being self-employed. This result may also

indicate that pro-entrepreneurial types are more likely to find one of

these professions attractive. This finding is in line with a previous study

on entrepreneurial personality and vocational choice by Sorgner (2012),

who shows that people with a pronounced entrepreneurial personality

are more likely to choose enterprising (e.g., manager, lawyer) and

artistic (musician, filmmaker) professions. Another effect that becomes

stronger but does not change its negative sign is for persons in

elementary occupations, who appear to be generally less

entrepreneurial.

An interesting result is obtained with regard to physical,

mathematical, and engineering professionals. According to the

univariate model, being in these professions had a significantly negative

effect on the propensity for self-employment; however, the results of the

bivariate probit model indicate that people who choose one of these

24

professional environments are more likely to be self-employed. This

result indicates that while these occupational environments appear to

attract people with pronounced entrepreneurial attitudes, they are less

conducive to the actual exercise of self-employment. The results for the

service workers should be interpreted cautiously since this was the only

group of professions for which our instrument variable does not have a

significant effect. The positive effect of being a service worker on the

probability of self-employment may indicate, for instance, that a rather

small minimum efficient size and low requirements with regard to

human and financial capital make entrepreneurial entry relatively easy

in this type of profession. However, based on our results we cannot

exclude the possibility that this effect is partly due to self-selection of

more entrepreneurial people into these professions.

Finally, as mentioned above, the results for health professionals,

teaching professionals, and craftspeople should be interpreted based

on the findings from the univariate probit regression, since self-

employment choice and the choice of one of these professions are less

likely to be interrelated. Therefore, we conclude (based on the results

from Table 5) that life science and health professionals, along with

trades workers, are more likely to become self-employed due to an

occupational environment that is conducive to entrepreneurship, rather

than due to entrepreneurial attitudes. Indeed, self-employment in both

these professional environments is highly regulated, meaning that

entrepreneurial opportunities in these occupations are highly

standardized and easy to pursue for those having the necessary human

capital.

Our study also sheds some light on the determinants of vocational

choice. Father’s occupation seems to have a strong effect on his child’s

vocational choice in almost all occupations, but other personal

characteristics are worth mention, too. In line with previous literature,

we find a significant relationship between personality traits and

vocational choice (Holland, 1985; Filer, 1986; Sorgner, 2012).

25

Interestingly, openness to experience—a personality dimension that is

often related to creativity and entrepreneurship—appears to be

negatively associated with the probability of being a service worker or

working in an elementary occupation. This might be why these

occupational groups demonstrate the lowest self-employment rates

(see Figure 2). Interestingly, people with a high willingness to take

risks—an attribute of many entrepreneurs (Caliendo, Fossen, and

Kritikos 2009)—are significantly less likely to be teachers, which can be

regarded as a relatively secure profession in terms of earning risk and

the risk of unemployment.

6. Conclusions

We have argued that self-employment is a dynamic process that occurs

in several steps. Specifically, we identified vocational choice as a first

step toward the decision to become self-employed. The choice of

profession occurs relatively early in life and tends to have a pronounced

effect on the career path, identity, and the accumulations of occupation-

specific human capital. Hence, to a certain degree, self-employment

may be occupation-specific in that certain occupations provide an

environment that is more conducive to self-employment than others. We

show that a person’s profession makes a considerable contribution to

explaining the decision to be self-employed. At least two reasons can

be identified for such an effect. First, people with pronounced

entrepreneurial attitudes may choose an occupation that complements

their values and matches their needs, which is then the reason for

observing relatively high self-employment rates in occupations such as

business professionals, legal professionals, and artists. In this case,

occupation-specific self-employment rates mainly result from the high

entrepreneurial potential of those choosing these occupations. Second,

certain occupational environments might be conducive to fostering self-

employment even among those who originally did not intend to found an

own business. In this case, working in a certain occupation might

26

stimulate the development of entrepreneurial skills (e.g., in managerial

occupations), or the occupation may be characterized by a high share

of standardized entrepreneurial role models or entry regulations that

make self-employment particularly attractive (e.g., for health

professionals, craftsmen). Hence, both the choice of a profession and

entrepreneurial choice appear to be highly interrelated and constitute an

important part of a person’s career development.

Our findings contribute to the literature on entrepreneurial choice in

several respects. First, we demonstrate the importance of vocational

choice and of occupation-specific environment in the decision to be self-

employed. Second, we enrich the discussion by showing that

personality plays an important role in the development of an

entrepreneurial career, probably most acutely early in the life course

when vocational choices are being made. The choice of a certain

profession, in turn, may be regarded as an important step either toward

or away from self-employment. Third, we present an empirical model

that allows analyzing the two related choices and identifying the main

links between them. Specifically, our empirical setting utilizes a bivariate

probit model that provides consistent estimates of the model of

entrepreneurial choice accounting for its interdependence with the

choice of a profession.

Having shown a strongly significant effect of the choice of a certain

profession on self-employment, the question remains as to what makes

a certain professional environment more conducive to entrepreneurship

than others. These may include such things as minimum efficient size,

entry regulation, occupation-specific labor market conditions, and

entrepreneurial career models, among others. Analysis of these

conditions is a promising avenue for future research. After having

discovered the elements that make an occupational environment

conducive to entrepreneurship, the next topic of investigation would be

how certain occupations can become more entrepreneurially conducive

and how to achieve a good match between entrepreneurial individuals

27

and occupations that foster their entrepreneurial talent. Since

entrepreneurial choices are likely to develop out of previous career

choices, including choice of profession, entrepreneurial careers can be

at least partly predicted, and thus efficiently guided, early in the life

course. Hence, a better match between people who show early talent or

interest in entrepreneurship and vocations that are conducive to

entrepreneurship or at the very least not impede the development of an

individual’s entrepreneurial capacities may be addressed in the course

of early career counseling of young people.

The findings of this study also contribute to the literature on

vocational choice. Specifically, we show that occupation following is still

a pronounced phenomenon in Germany, albeit one with a great deal of

variation across occupations. Occupation following is particularly strong

in trade occupations (about 47 percent) and it is almost absent in

service occupations (about 3.6 percent). This finding points to the role

of intergenerational transmission of skills, attitudes, and preferences in

vocational choice. Furthermore, we found support for the literature on

the role personality traits play in vocational choice (Holland, 1985).

Overall, we conclude that entrepreneurial choice evolves from an

individual’s personality, attitudes, previous career choices, and

experiences in a given context (Obschonka and Silbereisen, 2012).

Thus, an integrated approach that studies entrepreneurship from a life-

span perspective and accounts for the role of different contexts and

their interplay appears to be a promising one.

28

References

Bauernschuster, Stefan, Oliver Falck, and Stephan Heblich (2009): Training and Innovation. Journal of Human Capital, 3, 323-353.

Borghans, Lex, Angela Lee Duckworth, James J. Heckman, and Bas ter Weel (2008): The Economics and Psychology of Personality Traits. Journal of Human Resources, 4, 972–1059.

Bosma, Niels, Jolanda Hessels, Veronique Schutjens, Mirjam Van Praag, Ingrid Verheul (2012): Entrepreneurship and Role Models. Journal of Economic Psychology, 33, 410-424.

Caliendo, Marco, Frank M. Fossen, and Alexander Kritikos (2009): Risk Attitudes of Nascent Entrepreneurs—New Evidence from an Experimentally Validated Survey. Small Business Economics, 32, 153–167.

Costa, Paul T. Jr. and Robert R. McCrae (1992): Revised NEO Personality Inventory (NEO-PI-R) and NEO Five Factor Inventory (NEO-FFI) Professional Manual. Odessa, FL: PAR.

Dohmen, Thomas, Armin Falk, David Huffman, Uwe Sunde, Jürgen Schupp, and Gert G. Wagner (2011): Individual Risk Attitudes: Measurement, Determinants, and Behavioral Consequences. Journal of the European Economic Association, 9, 522-550.

Dunn, Thomas and Douglas Holtz-Eakin (2000): Financial Capital, Human Capital, and the Transition to Self-Employment: Evidence from Intergenerational Links. Journal of Labor Economics, 18(2), 282-305.

Fabbri, Daniele and Chiara Monfardini (2008): Style of practice and assortative mating: a recursive probit analysis of Caesarean section scheduling in Italy. Applied Economics, 40, 1411-1423.

Filer, Randall K. (1986): The Role of Personality and Tastes in Determining Occupational Structure. Industrial and Labor Relations Review, 39, 412–424.

Fritsch, Michael, Elisabeth Bublitz, and Alina Rusakova (2012): Berufstätigkeit und Entrepreneurial Choice – Welchen Einfluss hat die Berufstätigkeit auf die Entscheidung zur unternehmerischen Selbständigkeit? In Christoph Köhler and Alexandra Krause (eds.): Arbeit als Ware? Zur Theorie flexibler Arbeitsmärkte, Bielefeld: transcript.

Fritsch, Michael, Alexander Kritikos and Alina Rusakova (2012): Who Starts A Business and Who Is Self-employed in Germany? IZA Discussion Paper, #6326, 2012.

Gathmann, Cristina and Uta Schönberg (2010): How General is Human Capital? A Task-Based Approach. Journal of Labor Economics, 28, 1-49.

29

Gerlitz, Jean-Yves and Jürgen Schupp (2005): Zur Erhebung der Big-Five-basierten Persönlichkeitsmerkmale im SOEP, Research Notes. Berlin: German Institute for Economic Research (DIW).

Greene, William (1998): Gender Economics Courses in Liberal Arts Colleges: Further Results. The Journal of Economic Education, 29, 291-300.

Greene, William (2008): Econometric Analysis. 6th edition, Upper Saddle River: Pearson Prentice Hall.

Holland, John L. (1985): Making Vocational Choices: A Theory of Careers. Englewood Cliffs, NJ: Prentice-Hall.

International Labour Office (1990): ISCO-88, International Standard Classificaton of Occupations. Geneva: International Labor Office.

Jonsson, Jan O., David B. Grusky, Matthew Di Carlo, Reinhard Pollak, and Mary C. Brinton (2009): Microclass Mobility: Social Reproduction in Four Countries. American Journal of Sociology, 114, 977-1036.

Karlsson, Tobias and Maria Stanfors (2011): In the footsteps of their fathers? Occupational following among Swedish tobacco workers. Continuity and Change, 26, 45-68. doi:10.1017/S0268416010000330

Kassouf, Ana Lucia and Rodolfo Hoffmann (2006): Work-Related Injuries Involving Children and Adolescents: Application of a Recursive Bivariate Probit Model. Brazilian Review of Econometrics, 26, 105- 126.

Katz, Jerome A. (1992): A Psychosocial Cognitive Model of Employment Status Choice. Entrepreneurship: Theory and Practice, 17, 29-37.

Kihlstrom, Richard E. and Jean-Jacques Laffont (1979): A General Equilibrium Entrepreneurial Theory of Firm Formation Based on Risk Aversion. Journal of Political Economy, 87, 719–748.

Knight, Frank. H. (1921): Risk, Uncertainty and Profit. New York: Houghton Mifflin.

Laband, David N. and Bernard F. Lentz (1983): Like Father, Like Son: Toward an Economic Theory of Occupational Following. Southern Economic Journal, 50, 474-493.

Laband, David N. and Bernard F. Lentz (1989): Why so Many Children of Doctors Become Doctors: Nepotism versus Human Capital Transfers, Journal of Human Resources, 24, 396-413.

Lazear, Edward P. (2004): Balanced Skills and Entrepreneurship. American Economic Review, 94, 208-211.

Lazear, Edward P. (2005): Entrepreneurship. Journal of Labor Economics, 23, 649-680.

Lucas, Robert E., Jr. (1978): On the Size Distribution of Business Firms. Bell Journal of Economics, 9, 508-523.

Nanda, Ramana and Jesper B. Sørensen (2010): Workplace peers and entrepreneurship. Management Science, 56, 1116-1126.

30

Nedelkoska, Ljubica and Frank Neffke (2010): Human Capital Mismatches along the Career Path, Jena Economic Research Papers 051-2010, Friedrich Schiller University and Max Planck Institute of Economics Jena.

Obschonka, Martin, Rainer K. Silbereisen and Eva Schmitt-Rodermund (2010): Entrepreneurial Intention as Developmental Outcome. Journal of Vocational Behavior, 77, 63-72.

Obschonka, Martin, Rainer K. Silbereisen (2012): Entrepreneurship from a developmental science perspective. International Journal of Developmental Science, 6, 107–115.

Parker, Simon C. (2009): The Economics of Entrepreneurship. Cambridge: Cambridge University Press.

Rauch, Andreas and Michael Frese (2007): Let’s Put the Person Back into Entrepreneurship Research: A Meta-Analysis on the Relationship Between Business Owners’ Personality Traits, Business Creation, and Success. European Journal of Work and Organizational Psychology, 16, 353–385.

Rhine, Sherrie L.W., William H. Greene and Maude Toussaint-Comeau (2006): The Importance of Check-Cashing Businesses to the Unbanked: Racial/Ethnic Differences, The Review of Economics and Statistics, 88, 146–157.

Schmitt-Rodermund,Eva (2004): Pathways to Successful Entrepreneurship: Personality, Parenting, Entrepreneurial Competence, and Interests. Journal of Vocational Behavior, 65, 498–518.

Schmitt-Rodermund, Eva (2007): The Long Way to Entrepreneurship. Personality, Parenting, Early Interests and Competencies for Entrepreneurial Activity Among the Termites. In Rainer K. Silbereisen and Richard M. Lerner (eds.): Approaches to Positive Youth Development, Thousand Oaks: Sage, 205–224..

Schneider, Benjamin (1987): The People Make the Place. Personnel Psychology, 40, 437–453.

Sorgner, Alina (2012): A Physician With a Soul of a Cook? Entrepreneurial Personality Across Occupations. Jena Economic Research Papers 063 – 2012, Friedrich Schiller University and Max Planck Institute for Economics, Jena.

Sorgner, Alina and Michael Fritsch (2013): Occupational Choice and Self-Employment—Are They Related? Jena Economic Research Papers 001–2013, Friedrich Schiller University and Max Planck Institute for Economics, Jena.

Wagner, Gerd G., Joachim R. Frick, and Jürgen Schupp (2007): The German Socio-Economic Panel Study (SOEP): Scope, evolution and enhancements. Journal of Applied Social Science Studies, 127, 139-169.

31

Zhang, Zhen and Richard D. Arvey (2009): Rule breaking in adolescence and entrepreneurial status: An empirical investigation, Journal of Business Venturing, 24, 436–447.

Zhao, Hao and Scott E. Seibert (2006): The Big-Five Personality Dimensions and Entrepreneurial Status: A Meta-Analytical Review. Journal of Applied Psychology, 91, 259–271.

Zhao, Hao, Scott E. Seibert, and G.T. Lumpkin (2010): The Relationship of Personality to Entrepreneurial Intentions and Performance: A Meta-Analytic Review. Journal of Management, 36, 381-404.

32

Appendix: Tables

Table A1: Definition of variables

Variable Description

Self-employment Dummy = 1 if respondent was self-employed in 2009; else = 0

Entrepreneurial environment Agglomerations Areas with high population density Urbanized regions Areas with moderate population density Rural areas Areas with low population density

Human capital

Years of education Number of years the respondent has been in full-time education Years of unemployment Number of years unemployed

Social capital

Either parent has been self-employed

Dummy = 1 if either parents was self-employed when the respondent was 15 years old

Father’s occupation Father’s occupation (ISCO-88) when respondent was 15 years old

Married Dummy = 1 if respondent was married in 2009; else = 0

Socio-demographic characteristics Male Dummy = 1 if respondent is male; else = 0

German citizenship Dummy = 1 if respondent is German citizen; else = 0

Age Years of age

Personality characteristics

Openness to experience Mean score on the 7-point scales for: “I see myself as someone who has an active imagination” “I see myself as someone who is original and comes up with new ideas” “I see myself as someone who values artistic experiences”

Extraversion

Mean score on the 7-point scales for: “I see myself as someone who is communicative, talkative” “I see myself as someone who is outgoing, sociable” “I see myself as someone who is reserved” (reversed)

Conscientiousness Mean score on the 7-point scales for: “I see myself as someone who does a thorough job” “I see myself as someone who tends to be lazy” (reversed) “I see myself as someone who does the things effectively and efficiently”

Agreeableness Mean score on the following 7-point scales: “I see myself as someone who is somewhat rude to others” (reversed) “I see myself as someone who has a forgiving nature” “I see myself as someone who is considerate and kind to others”

Neuroticism Mean score on the 7-point scales for “I see myself as someone who worries a lot” “I see myself as someone who gets nervous easily” “I see myself as someone who is relaxed, handles stress well” (reversed)

Risk propensity An 11-point scale based on the question: “Are you generally a person who is fully prepared to take risks or do you try to avoid taking risks?” The value 0 means “risk averse” and the value 10 means “fully prepared to take risks.”

33

Table A2: Correlation matrix

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

1 Self-employed (yes = 1, no = 0) 1

2 Agglomerations 0.048 1

3 Urbanized regions -0.039 -0.769 1

4 Rural regions -0.016 -0.385 -0.295 1

5 Years of education 0.169 0.122 -0.082 -0.063 1

6 Years of unemployment -0.026 -0.072 0.019 0.081 -0.16 1

7 Either parent self-employed (yes = 1, no = 0)

0.111 0.045 -0.033 -0.02 0.102 -0.057 1

8 Married (yes = 1, no = 0) 0.057 0.004 -0.006 0.002 0.019 -0.074 0.009 1

9 Male (yes = 1, no = 0) 0.091 0.001 -0.001 -0.0002 0.023 -0.022 -0.007 0.031 1

10 German (yes = 1, no = 0) 0.012 -0.057 0.041 0.026 0.113 -0.07 -0.009 -0.056 0 1

11 Age 0.182 0.016 -0.019 0.004 0.06 0.02 0.031 0.315 0.042 0.069 1

12 Conscientiousness 0.031 -0.022 0.01 0.019 -0.071 -0.011 -0.018 0.058 -0.082 -0.021 0.109 1

13 Extraversion 0.056 0.021 -0.014 -0.011 -0.033 -0.034 0.012 -0.016 -0.144 -0.014 -0.059 0.158 1

14 Agreeableness 0.006 0.007 -0.004 -0.004 0.031 0.004 -0.015 -0.01 -0.149 -0.013 0.009 0.259 0.073 1

15 Openness 0.123 0.06 -0.037 -0.036 0.165 -0.041 0.049 -0.044 -0.088 0.001 0.04 0.11 0.331 0.137 1

16 Neuroticism -0.064 -0.022 0.019 0.006 -0.084 0.051 -0.011 0.006 -0.201 -0.015 -0.056 -0.099 -0.133 -0.102 -0.023 1

17 Willingness to take risks 0.082 0.027 -0.026 -0.003 0.031 -0.004 0.033 -0.075 0.178 -0.02 -0.12 -0.09 0.162 -0.134 0.154 -0.143

34

Table A3: Descriptive statistics

Variable: Mean Median Minimum Maximum Standard deviation

Self-employed (yes = 1, no = 0) 0.124 0 0 1 0.329

Agglomerations 0.501 1 0 1 0.500

Urbanized regions 0.370 0 0 1 0.483

Rural regions 0.129 0 0 1 0.335

Years of education 12.871 12 7 18 2.696

Years of unemployment 0.603 0 0 23.8 1.540

Either parent self-employed (yes = 1, no = 0)

0.103 0 0 1 0.303

Married (yes = 1, no = 0) 0.614 1 0 1 0.487

Male (yes = 1, no = 0) 0.525 1 0 1 0.499

German (yes = 1, no = 0) 0.954 1 0 1 0.210

Age 43.005 43 18 84 10.723

Openness 4.472 4.333 1 7 1.163

Conscientiousness 5.892 6 1.667 7 0.878

Extraversion 4.841 5 1 7 1.149

Agreeableness 5.260 5.333 1 7 0.974

Neuroticism 3.710 3.667 1 7 1.206

Willingness to take risks 4.104 4 0 10 2.076